Accepted to AJ Preprint typeset using LATEX style emulateapj v. 11/10/09
NEW MASER EMISSION FROM NONMETASTABLE AMMONIA IN NGC 7538. II. GREEN BANK TELESCOPE OBSERVATIONS INCLUDING WATER MASERS Ian M. Hoffman St. Paul’s School, Concord, NH 03301
arXiv:1110.2238v2 [astro-ph.GA] 26 Oct 2011
and Stella Seojin Kim1 St. Paul’s School, Concord, NH 03301 Accepted to AJ
ABSTRACT We present new maser emission from 14 NH3 (9,6) in NGC 7538. Our observations include the known spectral features near vLSR = −60 km s−1 and −57 km s−1 and several more features extending to −46 km s−1 . In three epochs of observation spanning two months we do not detect any variability in the ammonia masers, in contrast to the >10-fold variability observed in other 14 NH3 (9,6) masers in the Galaxy over comparable timescales. We also present observations of water masers in all three epochs for which emission is observed over the velocity range −105 km s−1 < vLSR < −4 km s−1 , including the highest velocity water emission yet observed from NGC 7538. Of the remarkable number of maser species in IRS 1, H2 O and, now, 14 NH3 are the only masers known to exhibit emission outside of the velocity range −62 km s−1 < vLSR < −51 km s−1 . However, we find no significant intensity or velocity correlations between the water emission and ammonia emission. We also present a non-detection in the most sensitive search to date toward any source for emission from the CC32 S and CC34 S molecules, indicating an age greater than ≈ 104 yr for IRS 1-3. We discuss these findings in the context of embedded stellar cores and recent models of the region. Subject headings: HII regions — ISM: individual (NGC 7538) — ISM: molecules — Masers — Radio lines: ISM 1. INTRODUCTION
The sources in NGC 7538 comprise a well studied complex of forming stars at a distance of 2.7 kpc (Moscadelli et al. 2009). Eleven infrared sources have been identified (Werner et al. 1979), three of which (IRS 1, 9, and 11) have associated water and hydroxyl maser emission (Kameya et al. 1990; Hutawarakorn & Cohen 2003) and one of which (IRS 1) hosts the largest single collection of observed maser species in the Galaxy (e.g., Galv´an-Madrid et al. 2010). In addition to illuminating IRS objects, the water masers are observed at several additional locations in the complex. Furthermore, the water masers are observed over a much wider range in velocity (≈ 100 km s−1 ) than all other maser species, which have only ever been observed over the limited range −62 km s−1 < vLSR < −51 km s−1 near the apparent systemic velocity of vLSR ≈ −58 km s−1 . Qiu et al. (2011), based on observations at submillimeter wavelengths, have recently reported several additional sources of continuum emission near IRS 1 in NGC 7538 that are described as embedded stellar cores (see also Akabane et al. 1992; Kawabe et al. 1992). Two of these sources, MM2 and MM3, have gas masses comparable to that of the central star in IRS 1 (≈ 20 M⊙) and have positions in common with water masers. The ongoing massive star formation near IRS 1 is described by Qiu et al. (2011) as the “Trapezium” of NGC 7538. In Figure 1 we show both the IRS 1-3 region and the larger complex including all other IRS objects.
[email protected] 1 present address: Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA 02139
Surcis et al. (2011) have formed the most recent model of IRS 1, incorporating new observations of water and methanol masers with other line (e.g., Qiu et al. 2011) and continuum (e.g., Sandell et al. 2009) observations in order to delineate an outflow, torus, and circumstellar disk. This model can be used to explain the line emission in the narrow velocity range −62 km s−1 < vLSR < −51 km s−1 but it does not address the details of the outflow or other sources presumably responsible for the “high”-velocity line emission outside of this narrow range, typified by the water masers. The (J, K) = (9, 6) 14 NH3 maser was discovered serendipitously by Madden et al. (1986) toward four of 17 Galactic star-forming regions surveyed: W51, W49, DR21(OH), and NGC 7538. In 2010, we observed the maser for the first time since its discovery, using the EVLA (Hoffman & Kim 2011, hereafter Paper I). We found (9,6) maser emission at new velocities, covering the range −60 km s−1 < vLSR < −56 km s−1 , associated in position with IRS 1, and consistent with the kinematics of the model of Surcis et al. (2011). Elsewhere in the Galaxy, nonmetastable (J > K) ammonia masers are known to be variable (W49: Madden et al. 1986; W51: Wilson, Henkel, & Johnston 1990). For example, Wilson & Henkel (1988) observed intensity variability of the maser in W51 by a factor of 20 over a timescale of 10 months. In a search for common variability of ammonia and water masers and for other new constraints on this well studied star-forming complex, we have undertaken new K-band radio observations of NGC 7538.
2
Hoffman & Kim
61 28 30 104 AU
IRS 2 declination (J2000)
25
20
IRS 3 15
MM3
MM2
10
05 23 13 47.0
IRS 1 46.0
45.0
44.0
43.0
right ascension (J2000)
Figure 1. (a) At left are images of the continuum emission from the IRS 1-3 region. The greyscale is the VLA ‘B’-array image at 4.8 GHz (6 cm) from Hoffman et al. (2003). The contours are the EVLA ‘DnC’-array image at 18.5 GHz (1.6 cm) from Hoffman & Kim (Paper I). The crosses mark the locations of the millimeter sources MM2 and MM3 from Qiu et al. (2011) that have coincident water maser emission, as discussed in §3.2. The position registration among these data sets is uncertain by approximately 500 milliarcseconds. The greyscale flux ranges from 5.5 to 56.5 mJy beam−1 . The contour levels are 15, 52, 104, and 150 mJy beam−1 . The beam for the 6-cm image is 1.4 × 1.1 arcseconds (shown at lower left). The beam for the 1.6-cm image is 4.6 × 1.9 arcseconds. The length of the line at the upper right represents an image scale of 104 AU for a distance to NGC 7538 of 2.65 kpc. (b) At right is a plot of the positions of the infrared sources (triangles, Werner et al. 1979) and water maser positions (circles) with filled symbols indicating associated water maser emission observed by Kameya et al. (1990). The triangle symbols have been omitted from IRS 1, 2, and 3 for clarity. The rectangle enclosing IRS 1-3 at the center is the boundary of the image in (a). Overlaid on the symbols is a plot of the beam response of the Green Bank Telescope for our observations. The dotted and dashed lines are sidelobe peaks in the Airy pattern. The first three 18-GHz sidelobes are shown with dotted lines and the first four 22-GHz sidelobes are shown with dashed lines. For these observations, the main beam was centered on IRS 1 and has a full width at half maximum of approximately 30 arcseconds, completely enclosing the imaged area shown in (a). The sidelobe levels are given in §2. IRS 7 and 8 are suggested to be unassociated with the NGC 7538 complex (e.g., Zheng et al. 2001). Table 1 Observed Transitions molecule 14 NH 3 CC34 S H2 O CC32 S
transition
νrest (MHz)
9, 6 21 − 10 61,6 − 52,3 21 − 10
18499.390 21930.476 22235.120 22344.030
Note. — The notation for the transitions is J, K for ammonia, JN for thioxoethenylidene, and JKa,Kc for water.
2. OBSERVATIONS
We observed three separate epochs using the Green Bank Telescope of the NRAO2 on 2010 November 24, 2010 December 8, and 2011 January 22. The November and December epochs each had an integration time on the science target of approximately 30 minutes while the January epoch had a target integration time of approximately 1 hour. We used the first-generation twobeam K-band receiver system for nodding observations between the beams using the “lower” (18-22 GHz) horns and polarizers (Srikanth 2009). The polarizers are approximately 20% less sensitive at 18.5 GHz and 22.3 GHz than at their peak at 20 GHz (S. Srikanth, 2011, private communication). Dual circular polarizations were recorded. We employed four spectral windows and three2 The National Radio Astronomy Observatory is a facility of the National Science Foundation operated under cooperative agreement by Associated Universities, Inc.
level sampling in the NRAO autocorrelation spectrometer (ACS) yielding 16384 channels across a 12.5-MHz bandwidth, corresponding to a usable velocity coverage of −140 km s−1 < vLSR < 30 km s−1 and a channel spacing of 0.010 km s−1 . The rest frequencies and molecular transitions used to center the four windows on vLSR = −60 km s−1 are summarized in Table 1. Amplitude calibration, focusing, and pointing were based on observations of 3C48, 3C123, and J2322+509. Typical rms noise levels in line-free channels were 23 mJy for the November and December epochs and 13 mJy for the January epoch. In Figure 1 is shown the observed region. In Figure 1b is shown a plot of the positions of the infrared (IRS) sources (Werner et al. 1979) and water masers (Kameya et al. 1990) in NGC 7538 overlaid with the beam and sidelobes of the GBT at both 18 GHz and 22 GHz. The relative amplitude of the first four 22-GHz sidelobes compared with the main beam are, respectively, −29.9 dB, −30.9 dB, −35.6 dB, and −38.0 dB, and for the first three 18-GHz sidelobes: −28.6 dB, −30.2 dB, and −33.0 dB (Srikanth 2009). There is negligible antenna sensitivity in the nulls of the beam pattern between the plotted sidelobe peaks. 3. RESULTS AND DISCUSSION 3.1. NH3 Ammonia maser emission was detected in each of the three epochs. In Figure 2 and Table 2 are shown the results from 2011 January 22. We do not detect any variability among any of the epochs. Aside from multiplica-
Nonmetastable Ammonia Masers in NGC 7538
3
Figure 2. (top) GBT spectrum of the 14 NH3 (9, 6) masers in NGC 7538 from 2011 January 22. (The 2010 November and December epochs have no significant differences from the epoch shown, as discussed in §3.1.) The fitted properties of the emission features are summarized in Table 2. (inset) A magnification of the three weakest (9, 6) emission features. (bottom) The thermal emission from the (1,1) transition observed using the GBT by Urquhart et al. (2011) as part of the RMS survey. Their (1,1), (2,2), and (3,3) observations are discussed in §3.1. Table 2 Fitted Spectral Properties of Ammonia Emission vLSR (km s−1 ) −60.16(2) −59.72(2) −56.93(1) −56.14(1) −55.78(1) −50.20(2) −48.85(1) −46.71(3)
S† (Jy) 0.94(2) 0.44(6) 0.49(1) 0.34(1) 0.44(1) 0.15(1) 0.057(2) 0.023(2)
∆vFWHM (km s−1 ) 0.64(2) 0.44(2) 1.16(2)a 0.55(1)a 2.84(1)a 0.58(1) 0.82(3) 0.65(7)
Paper I label ‘SE’ ‘SE’ ‘NW’ b b b b
Note. — The number in parentheses is the uncertainty in the final digit. † The absolute intensity calibration of the observations is precise within ≈ 10% while the higher precision quoted here is appropriate for relative comparisons among the features. a The fits to these features are likely unreliable due to blending of several narrower features. b The bandwidth of the EVLA observations presented in Paper I did not cover these velocities.
tive scaling (±10%) representative of typical calibration uncertainties, the spectra from the three epochs have no significant differences. Although the absolute calibration is uncertain at a ≈10% level, the relative magnitude of
the amplitude parameters listed in Table 2 are precise at a level of 1% or better. The magnitude and precision of the fitted line centers and widths are unaffected by amplitude scaling. The ammonia spectral features at vLSR ≈ −50, −49, and −47 km s−1 represent the first maser species in NGC 7538, besides water, to be observed having vLSR > −51 km s−1 . These weak features are not necessarily associated with IRS 1. At 18 GHz, the GBT has a −33.0dB sidelobe at the location of IRS 9 when pointed at IRS 1. For example, the 20-mJy signal in Figure 2 may be a 40-Jy ammonia maser in IRS 9. A large amplitude for an ammonia maser is not without precedent: Madden et al. (1986) and Pratap et al. (1991) find the (9,6) maser in W51 to have a flux density greater than 50 Jy. IRS 11 lies in a null of the 18-GHz beam pattern and likely does not contribute to the observed signal at 18.5 GHz. We are planning EVLA observations in order to precisely image the locations of all of the ammonia spectral features. Urquhart et al. (2011) have described observations of NGC 7538 IRS 1-3 as part of the RMS survey of over 1000 massive young stellar objects.3 The correspond3
bin/RMS/RMS DATABASE.cgi
http://www.ast.leeds.ac.uk/cgi-
4
Hoffman & Kim
ing MSX source name is G111.5423+00.7776. The survey includes GBT observations of the metastable 14 NH3 (1,1), (2,2), and (3,3) lines as well as the 61,6 − 52,3 water masers in the present study. The RMS observations of G111.5423+00.7776 were conducted approximately 12 months prior to the current observations. The telescope pointing position in the RMS survey is within an arcminute of the pointing position of the current data, meaning that the 22-GHz beam pattern in Figure 1b is an accurate description of the RMS sky response. G111.5423+00.7776 is one of only nine sources in their sample to exhibit 14 NH3 (1,1) emission at more than one velocity. They find at least five broad (≈ 3 km s−1 ) thermal emission lines at vLSR = −76.3, −64.0, −56.9, −48.9, and −37.2 km s−1 , consistent with the single-dish and interferometric spectra presented by Zheng et al. (2001). Their (1,1) data are shown in the bottom panel of our Figure 2. We note three important distinctions between the metastable and nonmetastable ammonia data sets: (1) there is no (1,1) emission at the velocity vLSR = −60 km s−1 of the strongest and longest-lived (9,6) masers, although there is (2,2) and (3,3) absorption, (2) in contrast, the new groups of (9,6) emission features centered at vLSR = −56.9 and −48.9 km s−1 each have corresponding (1,1) emission features, and (3) the thermal (1,1), (2,2), and (3,3) emission lines are much broader than the (9,6) lines, suggestive of a nonthermal origin for the (9,6) emission, including the low-amplitude (9,6) features having vLSR > −51 km s−1 .
Table 3 Fitted Spectral Properties of Water Emission vLSR (km s−1 ) −105.1(1) −104.4(1) −103.0(2) −96.7(3) −80.4(2) −78.0(2) −77.0(2) −75.4(3)a −74.5(1)a −74.1(2)a −73.5(2)a −72.8(2)a −72.2(2) −68.0(7) −67.0(3) −65.9(1) −64.8(1)a −63.0(1) −62.0(1) −60.36(4) −59.38(1) −58.53(4) −57.65(1) −56.9(2) −55.02(1) −54.47(2) −53.10(5) −51.7(1) −50.40(1) −47.95(5) −45.46(5) −44.0(1) −43.01(2) −42.45(1) −38.88(6) −34.84(1) −26.0(1) −21.6(1) −19.9(1) −12.0(1) −10.94(7) − 7.8(1) − 6.4(1) − 4.2(1)
SNov † (Jy)
SDec † (Jy)
SJan † (Jy)
0.14(1) 0.12(6)
